22 replies to this topic

[kvasan]

I am in the process of deciding the AOD pump specification for charging a liquid from a 200 lit barrel to a 7500 lit vessel. The feed port on the vessel is at a height of 4 m from floor level. The liquid is viscous (viscosity: 6500 cp) in nature and pipeline size employed for transfer is 1.5” (discharge side) 2” (suction side) with discharge pipeline length around 7 m.

 

The pump was mounted at a height of 1.5 m from the floor level. The pump which I have selected gives a much lower flow rate (5 PLM immediately after pump discharge side itself at an  air pressure of 6 ksc) which is far lower than my desired charging rate of 20 LPM.

 

Selected AOD pump details:

 

Max. Flow rate: 133 LPM; Max head: 85 m (in terms of water)

Diaphragm and ball valve MOC: PTFE

Max air pressure:  8 Ksc.

Feed and discharge port size: 1”.  

As per the characteristic curve provided by the manufacturer, the pump delivers around 90 LPM at head of 20 m. I am obtaining only 5LPM immediately after the pump and my dynamic suction head will be around max 10 m

 

I know that viscosity of the actual fluid employed will reduce the flow rate. But the pump discharge flow rate is less than 10 % of the rated flow rate. I am surprised at the extent of pump flow rate reduction.

My questions are:

 

1. What is the reason for the very low flow rate obtained?.

2. Will Teflon diaphragm employed and suction lift (presently 1.5 m static) affect the pump discharge flow rate?

3. Any suggestion  on selecting a pump of higher order or using the existing pump with any changes  

 

 

[latexman]

1. Not enough information to answer.  A dimensioned sketch with elevations and lengths and known operating pressures and temperatures would help.

2. Yes.  Teflon overlay on diaphragm always reduces capacity a little.

3. Can you warm the drum to raise temperature of the contents and lower viscosity?  We do that in a "hot box".  We try to limit it to 60 C so no one gets burned.

 

Have you done a fluid flow evaluation to confirm the observed flow or point to some issue?  Don't forget about acceleration losses on pulsating flow!!!  Plot the "system curve" on the pump curve.  Bigger pipe will probably help too, especially on the suction side.  More air pressure will help too.  Is there pressure on the 7500 liter tank?  Can that be reduced?  Is the 7500 liter tank about full when the drum as added?  Can it be added when the tankl is empty?  Can you move the operation to a higher level and go in the top head of the 7500 liter tank?

[breizh]

Hi,

Did you try to increase the air pressure ?

I guess you talked to the vendor , what are the comments ?

As suggested by latexman , a way to improve is to reduce the viscosity of the product . If you don't have a hot box , you may consider to use hot water pool , I was doing that way in Australia or heating blanket .

 

I've added a reference from a famous vendor.http://www.marcopump...II_Complete.pdf

 

Breizh

[chemks2012]

Hi kvasan,

I have checked against viscosity correction chart and it seems that if you wish to have flowrate of 20LPM, you should design your pump roughly for 50LPM.

I have checked the maximum possible flowrate at the air pressure of 6kgf/cm2,  with following data [given by you and assumed where required] and it seems that you should get about 9LPM of 6500cp fluid which equates about 25LPM of water flowrate.

• Density: 1000kg/m3
• Suction side: 1elbow and 1ball valve, 1.5m suction lift, 1.5m total pipe run
• Discharge side: 3elbows, 1ball valve, 3.5m discharge lift, 7m total pipe run

This means that, if you have more fittings than that I have assumed above, you may not even get 9LPM.

My suggestions would be

1. Viscosity is the real culprit here. As suggested by others, find out if you could reduce it by warming up
2. Your pump is capable of 8kgf/cm2, and you are using only 6kgf/cm2, find out if you used the full capacity using 8kgf/cm2
3. You may not like this but I am wondering if it is possible to increase the discharge pipe diameter?

PS: Feed and discharge port of 1” looks bit smaller to me anyways especially for a viscous fluid like this.

Hope that helps.

KS

Edited by chemks2012, 03 March 2014 - 04:23 AM.

[keerthivasan]

Thanks all for your suggestions. I am herewith attaching the approximate line diagram of my system and characteristic curve of the pump selected for your valuable suggestions:

 

Application:

The manufactures supplies the viscous liquid in 200 lit MS drums. Our application is to blend the liquid contained in individual drums after charging them in the blending vessels. Two blending vessels of capacity total capacity 7500 lit. and working capacity of 6000 lit are available for this purpose. I am planning to unload the material in the 200 lit MS drum to the blending vessel using AOD pump (Total charging quantity: 6000 lit).

Nature of liquid:

Highly viscous polymer liquid (6500 cp at 30 deg C).  The polymer is heat sensitive and I am afraid I shall be able to heat the product. Normally we handle at a maximum temperature of 40 deg C. So the option of heating the liquid to reduce the viscosity is not available to me.

Vessel:

The operating pressure and temperature in the blender are atmospheric conditions. Two vessels are available to cater our needs. The feeding port on the blender is at a height of  almost 4 m from the floor level.

Suction side:

- Corrugated Flexible piping of 2 m (max) was employed

-  Flexible Pipe Size: 2”

- Suction Lift: 1.5 m (max)

Discharge side:

• SS pipeline of size 1.5” (Schedule No: 40)
• Total piping length : 7 m (max)
• Pipe fittings: 1 Valve + 4 Elbows + 6 Flanges.  

Pump details:

• Max. capacity: 133 LPM ( Flooded inlet)
• Max head: 85 m
• Diaphragm MOC: PTFE
• Inlet/Outlet port size: 1”

    (Characteristic curve attached)

 

Answers for all queries:

 

1. There is no pressure in the tank while charging operation. The vessel was completely empty when the material charging was carried out.

2.  I would want to carry out the charging operation by placing the 200 lit MS drum on the floor level.   Otherwise I would have to lift many drums for charging the operating quantity of 6000 lit into the blender.

3. I will increase the air pressure to 8 ksc and post the results soon.

4. My installation of discharge line with all necessary pipe fittings is complete. So I do not want to modify my discharge pipeline at this stage

5.  I have experimentally calculated the flow rate by opening the flange joint that is immediately after the pump discharge (marked as ‘1’ in the diagram attached). The discharge head would be very little for the above case. It was found to be only 5-6 LPM.

  If I assume a total head of 20 m (very high estimate) up to that point ‘1’

As per the characteristic curve, I get a flow rate (at air pressure of 4.8 Ksc) of 90 LPM corresponding to the head of 20 m.

Then why am I getting such a very low flow rate with the actual liquid. (Or Am I missing any factor apart from viscosity?)

 

 

Any inputs based on the above data provided will be very helpful to me.  Looking forward to your suggestions..

 

Attached Files

•  Diagram.jpg   116.13KB   9 downloads
•  Pump Curve.pdf   132.46KB   44 downloads

[keerthivasan]

chemks2012:

 

Hi

 

Thanks for helping me out.

I have posted flow diagram and pump curve for the selected pump.

My  flexible suction line is of 2"  and discharge line of 1.5" diameter SS pipeline.

The inlet and outlet port sizes available in the pump is  1".

 

Can you please elaborate the calculations how did you obtain a flow rate of 9 LPM for my system/ 25 LPM for water.

What is the usual  capacity reduction factor to be considered while dealing with viscous liquids?

[latexman]

Is your corrugated pvc hose on the suction side collapsing under vacuum?

 

What's the rheology of this polymer liquid?  Newtonian, non-Newtonian, shear thinning, shear thickening, etc.?

Edited by latexman, 04 March 2014 - 04:26 PM.

[chemks2012]

keerthivasan,

 

Please see the attached viscosity correction factor chart [quite straight forward ] for AOD pumps which would give you an idea on flowrate at different values of viscosity.

 

I haven't run calculation based on your new info yet but few observations from data you gave.

 

1) Pump data states 133LPM is available when you have flooded suction which you don't have or rather you have suction lift. And again, do not forget this is given for water.

 

2) Pump curve also states that NPSH (required) is nearly 7 to 8 feet. You need to check if you have NPSH(available) is more than this. I believe this would be bit tricky in your case, as you may not have data available on pressure drop for your liquid [6500cP] though that corrugated 2" suction flexible.

 

I would say there is problem on the suction end too.

 

3) Check if you could increase the air flowrate along with air pressure which would help increasing the flowrate.

 

Hope, that helps.

KS

Attached Files

•  Viscosity Correction Chart for Air Operated Diaphragm Pumps.pdf   80.56KB   52 downloads

Edited by chemks2012, 04 March 2014 - 06:19 PM.

[breizh]

http://www.all-flo.c...ity-curves.html

 

Consider this document : Impact of viscosity on flow rate and general guidance about design .

 

Breizh

[chemks2012]

keerthivasan,

 

I am quite determined to help you here

 

Please ignore that 9LPM comment I made earlier as it doesn't apply anymore with your newly given info.

 

Please see the link below for pressure drop calcs link [you may have to assume some parameters if you don't have handy]  for your suction side flexible which will give you some idea on pressure drop through such flexible. See below are the results with some assumptions I made [dimensions for t, T etc] and it seems you have 0.28bar pressure loss through that flexible at 20l/min.

 

Checked calcs based on your new fitting info considering 0.28bar loss on suction side and also assumed vapour pressure of liquid as 20mbar and it's seems that you are just at the boundary line of NPSHr and that might be creating problem. Of course, if you have more vapour pressure than what I assumed, NPSH is the problem for sure for this pump.

 

Could you have pump connection made from the drum/IBC lower level [close to bottom] so that you can take the advantage of liquid head [flooded suction] and then all your problems probably go away?

 

http://www.pressure-....mobi/0105.html

 

 

Pressure Drop Online-Calculator

Calculation output Flow medium: Viscous Polymer at 30degC / liquid Volume flow: 20 l/min Weight density: 100 kg/m³ Dynamic Viscosity: 6500 cP     Element of pipe: Corrugated pipe Dimensions of element: Internal diameter of pipe D 50 mm   Corrugation height t: 5 mm   Corrug. distance T: 5 mm   Length of pipe L: 2 m         Velocity of flow: 0.17 m/s Reynolds number: 0 Velocity of flow 2: - Reynolds number 2: - Flow: laminar Absolute roughness:   Pipe friction number: 490.09 Resistance coefficient: 19603.54 Resist.coeff.branching pipe: - Press.drop branch.pipe: - Pressure drop: 282.49 mbar   0.28 bar

 

KS

[keerthivasan]

Latexman:

 

Hi

 

 

The hose is not collapsing under vacuum.

The liquid is a Newtonian fluid ..

[keerthivasan]

Hi KS

 

Many thanks for your effort in helping me out.

 

I need some clarification in the calculation of NPSHa:

 

There are two ports on the liquid barrel. One is the feed port of size around 2 “and the other is vent port. The vent port shall be open during charging.

Pressure on top of the liquid in the barrel is atmospheric (10.3 m of water which translates to 11.3 m of actual polymer fluid)

Static head loss is 1.5 m;

Dynamic head loss is 0.28 bar as per your calculation (3.17 m in terms of actual liquid)

Vapor   pressure: 20 mbar (I have assumed the same value as you have as the polymer liquid has a low vapor pressure. (0.22 m in terms of fluid)

 

 NPSHa = 11.3 – (1.5) – (3.17) – (0.22) = 6.41 (≈ 6)

 

However NPSHr is 2.5 m (from Pump curve)

 

If difference of 3.5 m between theoretical and available NPSH not sufficient?

Please correct me if I am wrong with my calculations above.

 

Further with respect providing pump connection from the barrel lower level, the product is supplied in standard 220 lit MS drums by the manufacturer with two ports (feed and vent) on the top of the drums with no ports on the bottom.  Also I am unable to increase the size of the suction pipe size due to the limitation of the feed port size on the barrel.

[chemmu]

Dear Vasan,

 

Good Day to you.

 

Please could you confirm that where the outlet of airline is connected?

Is there any obstruction in the air outlet (vent)?

Please also confirm the maximum withstanding viscosity of pump with the vendor

 

Regards,

Chem.M

[chemks2012]

keerthivasan,

 

I believe as long as you have NPSHa>NPSHr, your pump should work but again people keep margin between them and not sure what your pump manufacturer recommends.

 

Please note that most of the time for diaphragm pump, supplier specify suction lift and not NPSHr which is again not given in your case, which is bit surprising.

 

Your calculation for NPSHa is OK except you should consider atmospheric pressure of 10.33m [equivalent to that of water]. I would also consider minor entry loss [part of friction loss] but you can ignore as it would be negligible.

 

Please note that I personally never verified that hose calculator [the link I provided but it gives good indication on pressure drop anyways] and it could be possible that the frictional losses might be even higher than 0.28bar. however, can you check if you could use smooth hose [of course internally smooth – please check with hose supplier, they do smooth hose for viscous fluid]?

 

One of the reasons, I am very keen to help you here is - I have very bad experience of corrugated hose in the centrifugal pump suction line and viscous fluid. I designed a pump and client has ordered accordingly and pump wasn’t working properly [giving lower flowrate than what it is designed for]. It took ages me to find that client was using [without my knowledge] 3m long small diameter hose which he used to connect to the tanker during offloading fairly viscous liquid while my calculation were based on 1m, 80mm hose. Then we had to change the hose as well as trim the impeller diameter [as client was stubborn not to use VSD] as again viscosity given to us was wrong.

 

It seems that you have lots of limitations in terms of what you can do to make your pump work and I can’t think of anything else.

 

Did you try having higher pressure/flowrate of supply air?

 

KS

Edited by chemks2012, 06 March 2014 - 05:02 AM.

[keerthivasan]

chemmu:

 

Hi

The exit air comes through a muffler fitted in the pump.I am not so sure about obstruction in the air outlet. 

Need to check on that.

 

i was told that 1" AOD pumps can handle viscosity up to 22000 cp while mine is 7000 max.

 

Thanks for your input.

[keerthivasan]

chemks2012:

 

Dear KS

 

I was also bit surprised that suction lift was not mentioned.

I will try using a smooth hose pump meant for viscous liquids.

I will have to arrest leaks in my existing air line which gives  a max pressure of 6 Ksc only

 

I am finally left only with the option of  increasing the air pressure and using a smooth suction pipe with the existing pump.

Will post the results soon.

 

Thanks for your valuable inputs.

[breizh]

Hi,

 

I 've the strong feeling that the technology or the size of the pump are not adequate !  Did you talk to the polymers manufacturer(s) , he should point you in the right direction ( technology suitable versus material handling) ?

Did you supply your sketch to the pump vendor ? if yes what is the response ?

I don't know how often you need to do this operation per year but I'm sure your operators will have plenty of trouble handling this material using AOD .....and probably you will lose materials because unable to pump out of the drums.

 

I try to help you by sharing experience .

 

Breizh

 

You may consider to receive the material in containers , let say IBC , and install them on a platform which could help !

[keerthivasan]

Hi Breizh

 

 

I did talk to the chemical manufacturer. They are employing 1.5" AOD pump with max flow rate of around 350 LPM  as against 1" pump used by me and operating at an air pressure of 8 ksc. Not sure about their exact head requirements.

 

I am yet to get response from pump vendor. Further regarding IBC, manufacturer must agree to it. (difficult one).

 

I am bit confused as to go for AOD pump with higher rating or altogether use a screw pump/gear pump.

[keerthivasan]

The pump manufacturer claims a max suction lift (dry) of 6.1 m.

 

Does it mean my suction side losses (both static and dynamic) must be less than 6.1 m. ?

 

What does the term 'dry' suction lift refer to ?

 

I have seen some other manufacturers quote suction lift both 'dry' and 'wet'.

[latexman]

It means starting with no liquid in the suction piping, except for the pipe/hose that is sitting in the liquid.  The pump must evacuate the air and pull the liquid up to it's suction.

[keerthivasan]

It means starting with no liquid in the suction piping, except for the pipe/hose that is sitting in the liquid.  The pump must evacuate the air and pull the liquid up to it's suction.

Thanks

 

What is the significance of 6 m in this context.

[latexman]

It is the elevation difference between the pump suction and the liquid level in the suction vessel/tote/drum.  By saying "max suction lift (dry) of 6.1 m" the pump suction can be higher than the liquid level in the suction vessel/tote/drum by 6.1 m maximum.

[Dazzler]

Hi everyone, 

 

With regard to the NPSHa calculation check by keerthivasan ( on 06 Mar 2014 - 04:59 AM) I would like suggest that the term for “acceleration loss” has not been considered.  You can research what this term would be and how to estimate it, but for reciprocating pumps it is relevant to the calculation and may shed further light on issues with the suction arrangement.

 

Some other points:

 

A. With regard to corregated hosing, apart from the likely additional friction loss, the corregations can harbour residual fluid when not in use and this fluid cools and dries, so effectively lines the inside of the hose and may reduce the effective internal diameter of the hose.

 

B. One would need to be sure that the open end of the hose is not directly perpendicular to the base of the drum or it could hold itself to the bottom by suction, preventing much liquid entry.

 

C.  Is it possible the polymer liquid raw material settles slightly in storage, resulting in a thicker or viscous phase at the bottom where the hose inlet is.

 

Dazzler